基于SAA的混合演化算法及其应用研究-控制与决策.PDFVIP

第 15 卷第4 期 2000 年7 月 Vol. 15 No. 4 CON TROL A ND DECISI ON 　July 2000 SAA * 何　霆　马玉林 金　铮 ( 150001) () 　 提出将模拟退火、演化策略和局部搜索算法相结合的混合演化算法, 以解决函数优化与 合优 化问题。该算法克服了上述三种算法在应用中的不足, 并具有搜索效率高、性能稳定的特点。具体算例验 证了该算法的有效性。 模拟退火, 演化策略, 局部搜索算法, 混合演化算法 406.2 F - Research on SAA based New Hybrid Evolutionary Algorithm and Its Application H e Ting, M a Yulin Jin Zheng (Harbin Institute of Technology) ( eneral Computer Company of Heilongjiang Province) AbstractA novel hybrid evolutionary algorithm is proposed to solve the problems of function and com- binatorial optimization. T he algorithm is combined with local search algorithm and evolutionary strate- gy with simulated annealing algorithm. This algorithm overcomes previous defects of the three algo- rithms above and has the characteristics of high search efficiency and stability T wo examples show that. the algorithm is practical and effective. Key wordssimulated annealing, evolutionary strategy, local search algorithm, hybrid evolutionary al- gorithm 1引　　言 SAA(ES) (LS) , ( HEA) , , ( ) NP , 2算法简介 ,, ; 2. 1HEA , [ 1] maxf (x) , s. t.x S ( 1) , ( ) ,S ,f (x) ( ) , , x S , SAA NP